The development of research in the field of capital market has undoubtedly been crucial for compre-hending the role it plays in the macroeconomic and investment environment of a nation as a whole. But there has not been much research done on how the markets are interconnected. More importantly, rese-archers have rarely paid attention to the South Asian markets in the past. Only a small amount of research is available to understand how returns interact in these markets, as will be demonstrated in this paper in the later section, leaving investors and policy-makers in the major South Asian economies to prog-nosticate the underlying factors causing the return volatilities. Keeping these issues in mind, this paper aims to identify the nature of time series stationarity of these markets, investigate cointegration among them, and also determine if there is any causality moving from one market to the other (s) and vice versa.

It is a well-known phenomenon that the COVID and the war in Ukraine have affected the world economy in many dimensions (Ciravegna & Michailova, 2022). Ongoing quantitative tightening, the conse-quence of the aforementioned factors, in the deve-loped economies to tackle high inflation have put tremendous pressure on the emerging and develop-ing world making lives miserable for the general population (Canuto, 2022; Dulaney, 2022; Ha et al., 2022; The Editorial Board, WSJ, 2022). The pande-mic, the war and the ongoing macroeconomic pre-dicament, as mentioned above, have made it imper-ative to investigate what factors are affecting the capital flows and returns of different equity markets. It is only evident through previous research that the macroeconomic uncertainties and issues everely affect the equity markets of both developed and developing economies (Gunay & Can, 2022; Kara-nasos et al., 2022; Ma et al., 2022). Moreover, pre-vious research also identified that several economies are often interconnected and they may also change due to changes in the macroeconomic factors (Co-oray & Wickremasinghe, 2007; Smith et al., 1993). Hence, the recent macroeconomic developments of the world, which is a consequence of the develop-ments in the macroeconomic environment of the world following the global recession in 2008, the COVID – 19 pandemic, and the recent macro-econo-mic debacle as a consequence of the war in Ukraine, have made the author of this paper to lift the veil of opacity and reveal the intertwined relationships among the major South Asian markets of this region, namely India, Bangladesh, and Pakistan. 

An inquisitive mind may be interested in the reasons for choosing the South Asian region and only the THREE aforementioned economies, not the others. To reply to their potential queries, a rationale is being laid out here. First of all, India, Bangladesh, and Pakistan are USD 3.38 Trillion, USD 465 Bil-lion, and USD 383 Billion economies, respectively (Finance Division, Ministry of Finance, 2022; Khan, 2022; Natti, 2022). The world GDP right now is USD 95 Trillion (Dimitropoulou, 2022). Therefore, the aforementioned THREE South Asian economies collectively account for around 4.45% of world GDP. India alone accounts for around 3.56% of world GDP. Market Capitalization of the major stock exchanges of India, Bangladesh, and Pakistan are USD 3.402 Trillion, USD 55.65 Billion, and USD 32.84 Billion respectively (Dhaka Stock Exchange, 2022; Guagenti, 2022; Mumbai, PTI, 2022; PSX, 2021; PTI, 2022). The aforementioned data indicate that market capitalization to GDP ratios for India, Bangladesh, and Pakistan are 100.67%, 11.97%, and 8.57% respectively. These data indicate that the growth of the capital market relative to GDP in most of these economies are not good and may be indi-cative of economies with underdeveloped financial institutions (Khrawish et al., 2010). More import-antly, these markets are most likely to be under-valued (CFI, 2022). It is intuitive that one of the sources of this undervaluation may very likely be the interrelationships among the markets themselves. This is why, an investigation into this matter is extremely important. Moreover, total population of India, Bangladesh, and Pakistan as a ratio of the world population (as calculated by the author based on data from the world bank group) is around 22.72% (The World Bank Group, 2022b, 2022c, 2022a). Therefore, any research that ignores around one fourth of the world population seems to be in-complete. In addition, as will be discussed later, this region has hardly received any attention from the researchers of the world. However, the growth pot-ential of these regions is tremendous. Despite the ongoing macroeconomic predicament, different pro-jections by multilateral development banks have shown that India, Bangladesh, and Pakistan econo-mies are likely to grow at 6.7-8.2%, 6.6%, and 3.5% respectively in the coming year despite grim eco-nomic outlook for the world economy as evident by only a 3.2% growth projection in 2022 and 2.9% in 2023 (Asian Development Bank, 2022; Harshita Tyagi, 2022; Murali & Maiorano, 2022; TBS Report, 2022). Consequently, findings from research on any factor(s) or phenomenon affecting the stock market in this area are likely to result in new knowledge. In addition, economic integration among the economies of this region has been tested earlier only to find that, India has been consistently influencing the growth of this region (Jayaraman & Choong, 2012). However, this research was only carried out con-cerning a few relevant macro variables, not the inte-gration among the equity markets. Only one study was found on the topic of concern of this paper, which was also conducted prior to the global finan-cial crisis in 2008, the COVID impact, the on-going inflationary macroeconomic phenomenon, and a major stock market crash in Bangladesh (however, there are arguments about whether to call it a crash or a severe market correction), which brought about a plethora of changes including introduction of new indices (Saha, 2012). It is argued in that research paper that the stock markets of this region are not even efficient in the semi strong form, despite the fact that there is no causality among returns (Cooray & Wickremasinghe, 2007). Owing to these and the other motives listed above, the author developed a keen interest in investigating if there were any causal relationships between the daily returns of the key stock markets in India, Bangladesh, and Pakistan.

Causality tests have long been a well-liked technique for researchers looking for underlying the occur-rences of countless economic phenomena. Causality tests in the capital market and the macroeconomic domains have historically provided solutions to many unanswered "How" and "What" questions. Historically, different literature used different met-hods to establish relationships among key interest variables concerning market returns and other rele-vant macroeconomic variables. Smith et al. (1993), for example, used Granger causality on weekly returns from markets in the United States, the United Kingdom, West Germany, and Japan from 17 January 1979 to 26 June 1991 to find evidences of unidirectional causality running from the United States to the rest of the world immediately after the October 1987 global crash. They also found evi-dence of brief periods of causality running from other markets to the US (Smith et al., 1993). Another study employed linear and nonlinear Granger causal-ity tests to examine dynamic link between daily Dow Jones stock returns and percentage changes in trading volume on the New York Stock Exchanges only to find compelling evidence for a causal link between returns and volume that is nonlinear in both directions (Hiemstra & Jones, 1994). Rjoub et al. (2009) examined dynamic connections between stock returns and dividend yields in the American and Japanese equity markets using Granger-causality tests. They concluded that real returns cannot be caused by growth rates of dividends. The dividend growth rate does not Granger cause real returns, but the ratio of dividends to price (with and without lagged dividends) can predict returns (Rjoub et al., 2009).

Studies, such as the ones conducted by Ramasamy & Yeung (2005), find that investigating the links bet-ween exchange rates and stock markets has also sparked considerable interest. They, Ramasamy & Yeung (2005), used typical Granger causality app-roach on nine east Asian economies (Hong Kong, Malaysia, Singapore, Thailand, South Korea, the Philippines, Taiwan, Indonesia, and Japan) to exa-mine causal relationships. According to the duration of the chosen time, they discovered that the direction of causation frequently exhibits sot of a hit-and-run behavior suggesting that the Granger causality conclusions needs be interpreted with care(Rama-samy & Yeung, 2005). Another research used daily return data collected from 1995 to 2001 to invest-igate the connection between exchange rates and stock prices in Bangladesh, India, Pakistan, and Sri Lanka(Smyth & Nandha, 2003).This study by Smyth & Nandha, (2003) applied Engle-Granger two-step and the Johansen cointegration methods to conclude that no long-run equilibrium relationship is present between the two underlying financial interest vari-ables in any one of the four aforementioned coun-tries. Another study examines causal relationships between stock returns and exchange rate fluctuations for India, Indonesia, Japan, Korea, Malaysia, Philip-pines, Singapore, Taiwan, and Thailand from 1 January 1997 to 16 August 2010 using the Granger causality test in quantiles to find that there are vari-ous quantile causal relationships in different times and quantiles (Yang et al., 2014). A study by Ning et al. (2019) aimed to examine the link between Hong Kong and Shanghais overall market return, interest rate, money supply, exchange rate, and inflation rate using the APT, VECM, and Granger-Causality tests. They found that long- and short-term stock market returns in Hong Kong and Shanghai do have a link with the chosen macroeconomic factors (Ning et al., 2019; Alam, 2020).

Researchers also had interests in predicting causal relationship among other macro variables (other than the ones specified above), stock market phenomena and stock returns. For example, the industrial pro-duction index has been found to have a bidirectional causal relationship with the BSE Sensex, a signi-ficant Indian stock market index, according to a study by Singh, (2010) that used correlation, unit root stationarity tests, and Granger causality tests on monthly data between April 1995 and March 2009 (Singh, 2010). Zahid et al. (2022) used Granger causality test and ARCH-GARCH (1,1) models to conclude that environmental pricing, as exhibited by the energy market, and the stock markets in Bang-ladesh, Pakistan, and India have a bidirectional causal relationship (Zahid et al., 2022). Another research based on China used wavelet method to indicate that stock returns and investor sentiment have a strong bi-directional nonlinear causal link at different time frames (Chu et al., 2016).

Apart from the dimensions mentioned above, COVID impact on causality of stock returns has also been tested in different markets. Gherghina et al. (2020) tested the same in USA, Spain, Italy, France, Germany, UK, China, and Romania. It has been identified using auto regressive distributed lag (ARDL) and Granger causality that there existed no impact of Chinese COVID-19 records on the Roma-nian financial market, both in the short-term and the long-term (Gherghina et al., 2020). Another study using Granger causality and targeting Bangladesh, India, Pakistan, and Sri Lanka finds that there exists a unidirectional causality running from the number of confirmed COVID cases on market returns of the concerned markets (Rahman et al., 2022). 

However, number of research papers on the inter-relationship of market returns is really scarce. A study on the efficiency of the stock markets of India, Pakistan, Sri Lank, and Bangladesh used Granger causality only to find that these markets do not have any interrelationships (Cooray & Wickremasinghe, 2007). Another study by Valadkhani & Chancharat, (2008) examines whether there is cointegration and causality between the stock market price indices of Thailand and Australia, Hong Kong, Indonesia, Japan, Korea, Malaysia, the Philippines, Singapore, Taiwan, the UK, and the US using monthly data from December 1987 to December 2005.The study of Valadkhani & Chancharat, (2008) finds that there is no long-term association present between the stock prices of Thailand and these other nations. In the short term, however, Granger causalities running from Thailands stock market to Malaysia, Singa-pore, and Taiwan, as well as from Thailand to Hong Kong, the Philippines, the UK, Indonesia, and the US have been found. Bidirectional causalities have also been found among stock returns of Thailand, Malaysia, Singapore, and Taiwan (Valadkhani & Chancharat, 2008). Nonparametric test for Granger non causality and the conventional parametric Gran-ger non causality test have been used in another study to find that after the Asian financial crisis, the stock markets in Asia, including those in Hong Kong, Japan, Singapore, India, Malaysia, South Korea, and Taiwan, have grown increasingly glo-bally connected with those in Germany, the United Kingdom, and the United States (DeGooijer & Sivarajasingham, 2008). Another study used Granger causality on stock market indices from several Asian, European, and US markets to discover that Asian indices Granger-cause all other indices, and the same is true for US indices as well (Baumöhl & Výrost, 2011). 

A study by Dash, (2015) used ADF and Granger causality to imply that the Indian stock market is highly co-rrelated with the stock markets of South-East Asia and Japan, the Chinese stock market is influenced by the Indian market but not the others, and also the fact that the markets of India and Japan are integrated, whereas the Chinese market is more isolated (Dash, 2015).

These studies reviewed and listed above, however, fail to fill out gaps from several dimensions. First of all, all of these studies except one were conducted on either East Asian or on the developed nations leaving out major South Asian economies. India, however, has received some attention. Secondly, most of the studies were conducted to see the integration bet-ween stock returns and major macro variables. Inter-relations among the markets themselves were not studied. Thirdly, no research have been found, to the best of the knowledge of this author, to test the interconnections among the stock returns of Indian, Bangladeshi, and Pakistani stock markets following the global recession in 2008. Therefore, the impact of ongoing COVID crisis (which extends beyond TWO years by now), recent inflationary effects and the impacts of extended periods of monetary easing as well as short term tightening, interest rate hikes along with associated spillbacks (which are already translated into the stock prices) and stock market integration could not be captured by the prior rese-arch works. As a few studies cited earlier reveals, major macro phenomena are very much likely to affect integration among the stock markets(De Gooijer & Sivarajasingham, 2008; Smith et al., 1993). As a result, fresh research in this context is required. More importantly, this author could not ignore the rights of approximately one-fourth of the worlds population who have economies with enor-mous development potential and deprive them of the opportunity to learn about the interrelationships that may exist between the markets of their respective countries. Aside from that, this research will help the policymakers of all the THREE countries to develop effective macroeconomic policies for running their respective economies more efficiently, researchers in developing advanced findings based on this rese-archs findings, and investors develop and implement effective strategies to better understand and reap the benefits of their investments in this regions stock markets.

To test cointegration and causality, at first Augmen-ted Dicky-Fuller (ADF) unit root test will be con-ducted using the following equation:

∆y_t= α+βt+γy_(t-1)+∑▒〖δ_i ∆y_(t-1) 〗+ ε_(t-1)……………………………………………. (1)

In equation (1), y is the daily returns; model parameters are denoted by α,β,γ, and ε_t indicates the white noise error term (Alhussayen, 2022; Sam-man & Al-Jafari, 2015). For the purpose of the study, daily returns from BSE Sensex30, DSEX, and KSI indices for the period January 31, 2013, and August 31, 2022 were taken. Altogether 7,113 data points will be included in this analysis. Granger causality will be applied to see if returns of M Granger causes N by estimating the following equations (Cooray & Wickremasinghe, 2007; Gher-ghina et al., 2020)

R_(m,t)= α_0+α_1 R_(m,t-1)+〖…+α〗_p R_(m,t-p)+β_1 R_(n,t-1)+〖…+β〗_p R_(n,t-p)+ ε_t……………….. (2)

R_(n,t)= α_0+α_1 R_(n,t-1)+〖…+α〗_p R_(n,t-p)+β_1 R_(m,t-1)+〖…+β〗_p R_(m,t-p)+ u_t……….…….. (3)

For applying cointegration test, it is essential to see if the returns are a I(1) processes. If not, cointe-gration tests cannot be applied(International Burch University et al., 2022; Majumdar, 2021; Mohite & Bhandari, 2022). The ADF test will offer that infor-mation. Johansens cointegration test will be applied similar to Mohite & Bhandari, (2022) to test for cointegration among the given markets. STATA will be used to apply all of the aforementioned tests. 

Following table summarizes the results of the ADF test for all the return series of the THREE given markets:

Table 1: Summary of ADF test.

Table 2: Granger Causality.

* significant at 10%, ** significant at 5%, *** significant at 1%

As Table 1 shows, the returns of all the markets are stationary at level or I(0) processes. Therefore, coin-tegration tests cannot be performed. The results of Granger causality, as presented in Table 2, show that there are bidirectional causalities between returns of India and Pakistans stock markets. It implies that the returns of Indian stock market Granger Cause the returns of Pakistani stock market and vice versa. Therefore, events that are likely to have an impact on overall market in any one of these markets will likely to affect the other. As a result, anyone with an in-terest in the equity markets of either India or Pakistan must stay up to date on all developments in both markets (India and Pakistan) in order to effectively analyze their collective impact on the returns of the market of their choice.

This study tested causal links among the major South Asian markets, namely India, Bangladesh, and Pak-istan. Granger Causality and ADF have been used to test causality and unit roots of the daily return of all these stock markets. Results confirm that there are no causalities among the market returns of Bang-ladeshi and Indian stock markets. However, bidi-rectional causalities have been found between the market returns of Indian and Pakistani stock markets. These findings imply that investors in both of these markets should be cautious and well-informed about the other in order to predict how their returns will behave in the future. Policymakers in one country must monitor developments in the other in order to craft and implement prudent policies for the growth of their own equity market.

The authors would like to express gratitude to those people who are supporting this research study and helps providing the necessary information and conti-nuous support as well.

The author does not declare any conflict of interest.